JPH08331319A - Method of operating raster input scanner - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve image quality by aligning areas of original images which are exposed with respect to each primary photosensor array. SOLUTION: In the case of a color input scanner that obtains a digital signal from a full color original image, a sensor bar which has three parallel photosensor arrays that separately correspond to different primary colors is moved with respect to the original image in the direction of scanning. The timing for the exposure of the photosensor in each array is determined accurately in each scan cycle, so that the optical center of gravity in each exposed area of the original image for all the primary color photosensors are overlapped over one other.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to a color image sensor array, and more particularly, to each of a plurality of primary color filtered photosensors in each scan cycle with respect to the same location on the original image being scanned. The present invention relates to an "electronic shutter" device that exposes light to the background.

[0002]

BACKGROUND OF THE INVENTION Monochromatic image sensor arrays generally consist of a linear photodiode array that raster scans a document bearing an image and converts the tiny image area seen by each photodiode into image signal charge. After the integration period, the image signal is amplified and transferred to a shared output line or bus by a continuously energizing multiplexer transistor.

US Pat. No. 5,148,268 discloses a color image sensor array for recording full color original images as digital data. This patent describes a plurality of linear photosensor arrays arranged in parallel on one sensor bar, which sensor array moves in a scanning direction perpendicular to the arrangement direction with respect to the original image. To be made. The photosensors in each array
It comes with a filter of a certain primary color. As a sensor bar containing three photosensor arrays moves along the original image, portions of the original image area are exposed for each photosensor array. As each filtered photosensor array passes through an individual region in the original image, an individual photosensor in each array outputs a signal according to the different color separation of that region. Thus, the three linear photosensor arrays produce three separate sets of signals, each associated with one primary color.

[0004]

One practical problem with sensor bars having three linear photosensor arrays arranged in parallel is that for all scans of the three arrays, perhaps each of the three arrays is accurate. Is to not see the same area of the original image. The spacing of the primary color filtered photosensor arrays will cause a time delay in the signals output from each of the three arrays. Since the sensor bar is generally moving continuously with respect to the original image, it is expected that each of the three arrays will be exposed to a slightly different area within an individual scan line of the original image.

The present invention manipulates the integration time or "electronic shutter" of individual linear photosensor arrays within a full color sensor bar to more closely align the exposed area of the original image with the direction of the sensor bar. Aim to align. In this way, each primary color photosensor passing through an individual sub-region of the original image will see the same sub-region in the original image, thus improving image quality.

[0006]

SUMMARY OF THE INVENTION The present invention is a method of operating a raster input scanner having a sensor bar having a first linear photosensor array and a second linear photosensor array parallel to the first linear photosensor array. I will provide a. The sensor bar is moved relative to an object to be scanned in a scanning direction substantially perpendicular to the first linear array and the second linear array later. For each of a series of scanning cycles as the sensor bar moves, the integration of the image signal from the first linear array of photosensors occurs with a first exposure period. For each of a series of scanning cycles as the sensor bar moves, the integration of the image signals from the photosensors of the second linear array is calculated during the first exposure period with a second exposure period shorter than the first exposure period. Give rise to start.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT FIG. 6 shows a raster input scanner 102 using a sensor bar 10, which obtains image data from a document 103 placed on a platen 104. The sensor bar 10 is linearly arranged in the direction perpendicular to the paper surface of FIG. 6, and moves in the direction of the arrow 105 together with the lens 106 which is also linearly arranged. As shown in Figure 1,
Each photosite region 14 on the sensor bar 10 has photodiodes 14a, 14b, 14c that represent three primary colors.
Is included. Here, the photodiodes 14a, 14
Although b and 14c are shown and described, amorphous silicon or another type of photosite such as a transparent electrode MOS type photosite may be used. Further, here, row 16 of the photosite area 14
Although a one-dimensional sensor array having one is shown and described, it may be a two-dimensional sensor array using a plurality of rows of the photosite regions 14.

As shown in FIG. 2, each photosite region 14
Has a two-stage transfer circuit 20, and this transfer circuit 20 is
Together with the photosite region 14 and the buffer amplifier 33, a photosite cell 15 on the front side of the array is formed. In each cell 15, the image signal charge from the photosite region 14 is transferred by the circuit 20 to the amplifier 33, amplified by this amplifier 33 to the desired potential, and then transferred to the shared video output line or bus 22. Shift register
Logic circuit 24 provides a timing control signal PxSel for connecting each cell 15 to bus 22 in the correct timing sequence.

The sensor bar 10 can be used, for example, to raster scan a document. In that case, the sensor bar 1
0 is moved with respect to the original in the low-speed scanning (sub-scanning) direction perpendicular to the vertical axis of the sensor bar 10, that is, stepwise feeding. At the same time, the sensor bar 10 scans the document line by line in the high-speed scanning (main scanning) direction parallel to the vertical axis thereof. The image row being scanned is illuminated and the light reflected from the document is collected on the photodiode in the photosite region 14. During the integration period, a charge appears on each photodiode that is proportional to the reflectance of the image area seen by each photodiode. This image signal charge is then transferred to the two-stage transfer circuit 20.
Are sequentially transferred to the signal bus 22 via the amplifier 33 in a predetermined timing order.

The two-stage transfer circuit 20 of each cell 15 has a first-stage transfer transistor 26 for transferring the image signal charges from the photodiodes 14a, 14b, 14c to the amplifier 33.
a, 26b, 26c and the second stage transfer transistor 28
Is included. Transistors 26a, 26b, 26c are connected in series by lines 25a, 25b, 25c to one electrode of photodiodes 14a, 14b, 14c and the input gate of amplifier 33. Photodiode 14
The other electrodes of a, 14b, and 14c are grounded. The baice charge injection transistor 36 injects a bias charge into the line 25 at the intermediate node 37, for example a charge which electrically obtains a potential V _FZ of Fat Zero. The reset transistor 38 includes the transistor 28 and the amplifier 3
The reset voltage V _R is controlled by the node 39 during the period 3.

A suitable clock 45 provides the appropriate source pulses ΦR, ΦTA, ΦTB, ΦTC, ΦT2, ΦFZ and also V _FZ . Pulse ΦTA, ΦTB, ΦTC, ΦFZ,
V _FZ is provided to inject bias charge into lines 25a, 25b, 25c, and pulses ΦT2, ΦR are provided to set node 39 to voltage V _R. Pulses ΦTA, ΦTB, and ΦTC of different amplitude are photodiodes 1
It is provided to transfer the image signal charge from 4a, 14b and 14c to the amplifier 33. The PxSel signal used to multiplex the charge amplified by amplifier 33 and output to shared video output bus 22 is provided by shift register and logic circuit 24.

In operation, pulse .PHI.R energizes reset transistor 38 and applies reset voltage VR to node 39 to reset the input to amplifier 33. Then pulse Φ
TA and ΦT2 respectively activate the transistors 26a and 28 of the two-stage transfer circuit 20, and transfer the image signal charge accumulated in the photodiode 14a of each cell 15 to the amplifier 33. To facilitate the transfer of the image signal charge, the amplitude V (ΦTA) of the pulse ΦTA is the amplitude V (ΦT2) of the pulse ΦT2.
Less than. During injection of bias charge, pulses ΦTA, Φ
FZ goes "high", while pulse V _FZ goes "low". After a preset period of time, pulses ΦTA, ΦFZ,
V _FZ returns to their nominal level.

As the sensor bar moves relative to the original document being scanned or other object to be scanned, each primary color photodiode 14a-14c in a cell 15 on the sensor bar is sequentially exposed to its respective location on the original document. . When each photodiode 14a-14c is exposed to an individual location on the original, each photodiode accumulates a charge proportional to the amount of its primary color within a portion of the original image, so that three charge signals (each signal is an individual signal). (Corresponding to a certain primary color in the place) occurs. These three primary color signals are supplied to the photodiodes 14a ...
Clock 45 activates various transistors in transfer circuit 20 to read out from 14c onto video output bus 22 in a usable order.

In connection with the present invention, the important parameters ultimately controlled by the action of the clock 45 include the primary color photosensors 14a-14 in each cell on the sensor bar.
An integral of the charge accumulated on c is included. The integration of these charges ensures that charges of a particular magnitude (which are directly related to the amount of primary color for each location within the original image being scanned) will be transferred to the video output bus 22 through the transfer circuit 20. Shows. In short, each photodiode 14
It is possible for a to 14c to receive light for charge generation at any time. An important parameter for controlling the individual photodiodes during scanning is the timing of the event that the individual photodiodes are discharged (finally producing a video image signal). As the sensor bar moves with respect to the image being scanned, each photosensor for each primary color has a fixed length of exposure time during which the individual photosensors are scanning the image being scanned. You can receive light from individual locations. The principle is the same as opening the shutter to expose the film in the camera.

For each scan cycle, the photodiode is temporarily shut off so that the charge accumulated during the scan cycle can be read out via the transfer circuit 20. As used in the claims, "integral" means the portion of each scan cycle in which light from the image or object being scanned can charge individual photodiodes. The length of time during each scan cycle during which this integration can be performed is called the "exposure period". The longer the exposure period, the more light the light can produce on each individual photodiode. In the embodiment of the present invention, fat zero (fat-
zero) In the case of a bias charge injection device, the start and end of each exposure period are determined by the action of the transfer circuit 20. Each exposure period begins when the fat zero bias charge is injected into an individual photodiode and ends when the corresponding transistor 26 discharges that photodiode and transfers the charge through transfer circuit 20 to reset node 39.

FIG. 3 is a diagram showing the relative positions of areas on a hard copy image scanned by a set of photodiodes 14a-14c in a conventional apparatus. As shown, the three photodiodes 14a-14c are represented by squares corresponding to the relative size and spacing of the light sensitive areas associated with the photodiodes found in a typical sensor bar design. Each photodiode 14a-
14c is assigned one of the three primary colors and is filtered so that the photodiode reacts only to that primary color. That is, the photodiode 14a has a blue filter, the photodiode 14b has a green filter, and the photodiode 14c has a red filter. In addition, FIG. 3 shows three columns corresponding to regions along the scanning direction (downward in FIG. 3) of the original image scanned by the individual photodiodes over time. Although the regions corresponding to different colors are shown as separate columns, in practice the three columns are effectively overlapping and follow the same path for the image or object being scanned. However, in FIG. 3, the behavior of the three photodiodes corresponding to blue, green and red is shown as separate columns for convenience of description.

In FIG. 3, as the three photodiodes move continuously downwards to scan the original image, each thick horizontal line in each column represents the timing at which the exposure period begins for the individual photodiode. An "X" drawn just below each thick horizontal line indicates the extent of the image area scanned by the particular photodiode that begins the exposure period at the thick horizontal line. The black dot at the center of each X in FIG. 3 represents the center or centroid of a particular small area of the image scanned during each exposure period. This is an important concept when exposing multiple overlapping areas of the original image with different photodiodes. Each X
The fact that the area covered by is larger than the area corresponding to the individual photodiodes is
Due to the fact that the photodiodes are "on" (exposing an area of the original image being scanned) for a particular exposure period, while 4a-14c move continuously with respect to the image being scanned. .

The dashed horizontal line in FIG. 3 indicates the relative position of the area being scanned. The spacing between the dashed horizontal lines is assumed here to represent 1/6 of the scan line corresponding to the scanning cycle of the sensor bar. As can be seen from the figure, each of the photodiodes represented by the photodiodes 14a to 14c vertically surrounds all the scanning lines (six horizontal horizontal lines), and the boundary line of the photodiode is from the boundary line of another photodiode to the scanning line. 1/2, that is, three broken horizontal lines are provided. This particular spacing is the normal spacing required to produce a photosensor on an integrated circuit.

Three corresponding to each of the three primary colors
For each cell in the sensor bar having a number of photodiodes 14a-14c, there must be a routine to read each photodiode in a consistent order. In the conventional device, photodiodes 14a-14
Techniques for reading out c in a simple number sequence are known. That is, first, a photodiode, for example 14a,
Allowed to emit that signal onto the video bus 22,
Next, the photodiode 14b and the photodiode 14c are read. For each scan cycle, the individual photodiodes are assigned the same length of integration time. However, from FIG. 3, in the case of the conventional numbered same exposure period technique,
In an actual device, it can be seen that the center of gravity of the image scanned downward by the continuous primary color photodiodes is skewed (shifted) in the drawing. From the thick horizontal line indicating the beginning of each scan and the number indicating the order of the exposure period of each photodiode attached to each horizontal line, the center of gravity of the image portion being scanned is not in a straight line (the three columns are effectively the original lines). Overlapping on the image). For example, the blue photodiode 14
a is always scanning a location on the original image that has a center of gravity just below that of the second row of green photodiodes and the third row of red photodiodes. This skew of the center of gravity for different primary colors can have undesirable effects when scanning complex primary images such as photographs.

The independently controllable exposure periods of the different primary color photosensor arrays of the present invention allow the timing and time of exposure of each primary color photodiode array to be finely and precisely adjusted so that different primary color photodiodes can be scanned. An apparatus is provided which allows the center points or centroids of existing image portions to be accurately aligned, ie, superposed.

FIG. 4 shows a device in which the centroids of the regions scanned by each primary color photodiode are perfectly aligned (the three columns shown are actually overlapping, so their centroids are strictly on the original image). In the same place). In order to thus align the centroids of the image areas being scanned by each primary color photodiode (compared to the X formed by the image areas being scanned by different photodiodes), the sensor bar for the image being scanned. The blue photodiode 14a, which is the leading edge of the, is assigned a longer exposure period than the green photodiode 14b. The red photodiode 14c is assigned a much shorter exposure period than the green photodiode 14b. Thus, as a result of gradually shortening the exposure period of the green photodiode 14b and the red photodiode 14c, the green photodiode 14b and the red photodiode 14c scan a region slightly smaller than the photodiode 14a forming the front edge of the sensor bar.

As a general effect, the purpose of progressively shortening the exposure period of the subsequent green and red photodiodes is to have adjacent photodiodes represented by three spaces made up of the horizontal line between the photodiodes 14a-14c in the scanning direction. It is to compensate for the dead space between them. In addition, it will be noted that the exposure period of each of the subsequent photodiodes will start later and end earlier than the blue photodiode in order to have the exact center or center of gravity of the image location scanned by each photodiode. Let's do it. Since the integration of the green photodiode 14b has a shorter exposure period than the blue photodiode 14a,
The centers of the areas being scanned are placed at the same center point along the scanning direction.

To achieve the desired relative shortening of the photosensor exposure period from the front to the rear of the sensor bar, the clock 45 controlling the bias injection and discharge of the individual photodiodes 14a-14c is adjusted. , The exposure period can be adjusted quite accurately. FIG. 5 shows a timing diagram for the exposure period of the primary color photodiode which realizes the alignment of the centroids of the scanning image shown in FIG.
In the embodiment of the present invention using fat zero bias (charge) injection, the exposure period is determined to begin with the fat zero bias (charge) injection and end with the discharge of the photodiode. The figure can also be applied to other devices for controlling photodiodes. The exposure time for an individual photodiode is given by the "thick" part of each line in the timing diagram. Reference characters N and N + in each integration period shown in FIG.
1, N + 2 ,. ． ． Correspond to different parts of the image being scanned. At the bottom of FIG. 5, the transfer circuit, for example, 20
The output on video line 22 is shown, showing the signal output by the set of three primary color photodiodes when the image can be read by.

Examination of FIG. 5 shows that the leading edge of the blue (B) photodiode is given a relatively long integration period, and is essentially immediately followed by another integration period. For each scan cycle represented by the long integration time of the blue photodiode, the exposure period of the second photodiode of the sensor bar (here, the green (G) photodiode) is about 2 / half of the exposure period of the blue photodiode.
It is intended to begin at point 3. The exposure period of this green photodiode is 2 times the exposure period of the blue photodiode.
/ 3 period continues. Finally, the exposure period given to the last photosensor (here, the red (R) photodiode) of the sensor bar is one-third the period of the blue photodiode on the front side, and the exposure is the exposure period of the green photodiode. Will start when is over. Therefore, the start of each exposure period is shifted by 2/3 of the longest exposure period for all of the three primary color photodiodes, and the relative ratio of the exposure period is
The exposure time is gradually shortened to become 1: 2/3: 1/3. (Returning to FIG. 4, the relative lengths of the scanning regions of the three photodiodes are 12: 10: 8 in terms of the number of horizontal lines.
The reason why it is not 1: 2/3: 1/3 is that there is the length of the photodiode in the scanning direction and the interval between the diodes. )

[0025]

Focusing on the video line which represents the output on the video line 22 when the image signal is read from the photodiode, a regular pattern of the video signal despite the different exposure periods of the individual photodiodes. You can see that is output. Also, the readout signal on the video line essentially immediately follows the end of the exposure period of the individual photodiode.

The fact that the primary color photodiodes have considerably different exposure periods even when scanning the same primary image does not necessarily have a detrimental effect on the quality of the image signal read from the device. Not exclusively. On the contrary, it has been found that in one embodiment of the invention, the longer exposure time of the blue photodiode adequately compensates for the low light transmission of the filter used. Differences in signal values caused by different exposure periods for different primaries can be corrected in the image processing device, at least downstream of the photosensor.

[Brief description of drawings]

FIG. 1 is a plan view of a photosite region of a sensor bar.

FIG. 2 is a circuit diagram showing a pixel cell.

FIG. 3 is a diagram showing the behavior of each photodiode with respect to the surface of the image being scanned and a plan view of a typical photodiode in a pixel cell, according to a conventional technique of reading a signal from the photodiode.

4 is a plan view, similar to FIG. 3, of a portion of the area being scanned showing an implementation of a readout method for a photodiode, according to one embodiment of the present invention.

5 is a timing diagram showing the operation of different primary color photodiodes and associated video lines for implementing the readout method shown in FIG. 4;

FIG. 6 shows a schematic diagram of a raster input scanner using a sensor bar.

[Explanation of symbols]

 10 Sensor Bar 14 Photosite Area 14a to 14c Photodiode 15 Pixel Cell 20 Two-stage Transfer Circuit 22 Shared Video Output Line 24 Shift Register / Logic Circuit 26a to 26c First Stage Transfer Transistor 28 Second Stage Transfer Transistor 33 Amplifier 36 Bias Charge injection transistor 36 37 Intermediate node 38 Reset transistor 39 Node 45 Clock

Claims (3)

[Claims] 1. A first linear photosensor array responsive to a first color and a second linear photosensor array parallel to the first linear photosensor array and responsive to a second color. A method of operating a raster input scanner with a sensor bar having a first linear photosensor array followed by a second linear photosensor array, the sensor bar being substantially orthogonal to the linear photosensor array. Moving in the scanning direction with respect to the object to be scanned, for each of a series of scanning cycles during the movement of the sensor bar, causing an integration of the image signals from the photosensors of the first linear photosensor array, Each having a first exposure period, and for each of a series of scanning cycles during movement of the sensor bar, the second line Form an integration of the image signals from the photosensors of the second photosensor array, the integration of the second linear photosensor array having a second exposure period that is shorter than the first exposure period, the integration being the first exposure period. A method which is started during an exposure period. 2. The method of claim 1, further comprising injecting a bias into each photosensor of the first linear photosensor array at each scan cycle, and at each scan cycle the second linear photosensor. A method comprising injecting a bias into each photosensor of the sensor array. 3. A first linear photosensor array and a second array arranged in parallel with each other, each responding to a different primary color.
A method of operating a raster input scanner comprising a sensor bar having a linear photosensor array and a third linear photosensor array, the sensor bar being scanned in a scan direction substantially perpendicular to the linear photosensor array. And with respect to each of a series of scanning cycles as the sensor bar moves, causing integration of an image signal from a photosensor of the first linear photosensor array, each of the integrations causing a first exposure period. For each of a series of scan cycles as the sensor bar moves, causing an integration of an image signal from a photosensor of the second linear photosensor array, the integration resulting in the first exposure period. Has a second exposure period lasting 2/3 of the second exposure period, the integration starting at 2/3 of the first exposure period. , For each of a series of scanning cycles as the sensor bar moves, causing an integration of an image signal from a photosensor of the third linear photosensor array, the integration being one third of the first exposure period. A method having a third exposure period lasting a period of time, the integration starting after the second exposure period.